TECHNIQUES FOR SEARCHING WEB FORUMS

• US 20080215541A1
• Filed: 03/01/2007
• Published: 09/04/2008
• Est. Priority Date: 03/01/2007
• Status: Abandoned Application

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Li, Hang, Xu, Gu
• Application Number:
  US11/681,165
• Publication Number:
  US 20080215541A1
• Time in Patent Office: Days
• US Class Current: 707/3
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A search system provides various techniques for facilitating information retrieval. The search system may identify alternate queries for an initial query submitted by a user to a search system. Upon receiving the initial query, the search system identifies questions that are related to the initial query and presents to the user the related questions as alternate queries. The search system may also identify messages within a discussion thread that include answers. The search system may also identify an expert relating to the subject of a query by searching through expert profiles containing keywords of discussion threads in which the expert participated.

Claims

1. A computer system for identifying alternate queries for an initial query, the system comprising:

a component that receives an initial query;

a question store containing questions;

a component that identifies from the question store questions that are related to the initial query; and

a component that indicates that the related questions are potentially alternate queries for the initial query.

2. The computer system of claim 1 including a component that outputs the alternate queries to a user and receives a selection of an alternate query as a final query.

3. The computer system of claim 2 including a component that submits to a search engine the final query, receives from the search engine search results, and outputs the search results to the user.

4. The computer system of claim 1 wherein the questions of the question store are questions of a discussion thread.

5. The computer system of claim 1 wherein the questions of the question store are queries submitted to a web page search engine.

6. The computer system of claim 1 including a component that identifies expanded queries of the initial query and wherein the component that identifies questions identifies questions that are related to each expanded query.

7. The computer system of claim 1 wherein the related questions are ranked based on their similarity to the initial query.

8. The computer system of claim 1 wherein the component that identifies questions selects questions based on a language model-based probability of a corpus of documents relating to the questions.

9. The computer system of claim 8 wherein the questions are queries submitted to a web-based search engine and the corpus includes content identified by search results of the queries.

10. A computer system for identifying an answer within a discussion thread, the computer system comprising:

a component that identifies candidate messages of the discussion thread submitted by persons other than an initiator of the discussion thread;

a component that removes as candidate messages those messages that are identified as likely not being an answer; and

a component that ranks the candidate messages remaining after the removal based on likelihood of being an answer.

11. The computer system of claim 10 including a component that receives a query, searches for discussion threads related to the query, and returns, as search results of the query, messages of the related discussion threads identified as high-ranking answers.

12. The computer system of claim 10 including:

a component that identifies messages of the discussion thread submitted by the initiator of the discussion thread; and

a component that identifies an initiator message as an answer

wherein the component that ranks the candidate messages ranks the initiator message identified as an answer along with the candidate messages.

13. The computer system of claim 12 wherein the initiator message identified as an answer is ranked higher than the candidate messages.

14. The computer system of claim 10 including a component that receives an initial query, identifies from a question store questions that are related to the initial query, outputs to a user the related questions as alternate queries to the initial query, receives a selection of an alternate query as a final query, submits to a search engine the final query, receives from the search engine documents as search results, and outputs to the user messages of the search result.

15. The computer system of claim 10 including a component that identifies messages that are not answers using a support vector machine classifier.

16. The computer system of claim 10 wherein the component that ranks candidate messages ranks using a ranking support vector machine.

17. A computer system for identifying an expert relating to a query, the computer system comprising:

a discussion thread store;

a component that identifies experts from among persons who are participants in discussion threads of the discussion thread store;

a component that creates, for each identified expert, an expert profile having keywords derived from messages of the discussion threads in which the identified expert participated; and

a component that receives a query, searches for an expert profile that matches the query, and indicates that the expert of a matching expert profile is an expert relating to the query.

18. The computer system of claim 17 wherein keywords are identified using a term frequency by inverse document frequency metric.

19. The computer system of claim 17 wherein the component that identifies experts identifies experts based on number of messages submitted by the expert.

20. The computer system of claim 17 wherein the component that searches for an expert profile searches based on keywords of the query.